Selective publication of research into antidepressants:consequences for the practice guideline 'depression'

It was recently shown in a publication in The New England Journal of Medicine that the publicly accessible information about the effectiveness of antidepressive medication is incomplete and biased. The authors calculated from published and unpublished data that the effectiveness of antidepressants as published in the literature is overestimated probably by 32%. Based on this publication and other publications questioning the effectiveness of antidepressants, adjustment of the Dutch practice guideline for depression is proposed: only severely depressed patients should be treated with antidepressants. Some measures to prevent publication bias are also proposed. These require the cooperation of all parties participating in registration ofdrugs and publication of drug research.